^{page 119 note 1} Danielou, Alain, Northern Indian music, I, London, Christopher Johnson, 1949, 46.Google Scholar

^{page 119 note 2} ibid., 34.

^{page 119 note 3} ibid., 56–7.

^{page 123 note 1} The intervals in cents are accurate only to the nearest whole figure.

^{page 125 note 1} Even the idea that there is a steady note intended is to some extent subjective. The oseillograms generally show a fairly steady stretch usually in the middle of the note filmed. There are exceptions, notably in Bundu Khan's recordings, where the tones are generally oscillating, and in one or two instances of Pannalal Ghosh's recording, where the pitch is gradually rising. If, however, no particular steady note is intended, then the objects of the analysis are immediately resolved to a large extent, and no fixed intervals can be established.

^{page 126 note 1} This phenomenon was also noted on the strobocon, an instrument consisting of 12 stroboscopic discs whose speed of rotation is variable. These 12 discs are adjusted to the perfectly tuned 12 semitones of the Western scale, so that when one of these tones ia fed into the strobocon, the corresponding disc appears to stop moving. The range of possible tones between the 12 semitones can be measured by slightly increasing or decreasing the speed at which these discs rotate. When a note on the sitar was fed into this instrument the markings on the disc in tune, first shifted rapidly in a clockwise direction (indicating a higher pitch), gradually slowing down (indicating a dropping pitch) before becoming more or less stationary. To determine the drop in pitch from the moment of impact to the steadying of the tone, the speed of the corresponding disc was increased to the point where the disc no longer turned in a clockwise direction at the moment of impact. The results were variable, depending on the force of impact, string length, and string tension, but indicated that 20 cents was by no means an unreal figure.

^{page 127 note 2} Alain Daniélou, op. cit., it, London, Halcyon Press, 1954, 142.

^{page 127 note 1} This is in accord with Daniélou's statements quoted on p. 119 of this article.

^{page 128 note 1} Some of these texts also mention the gāndhāra grāma which is said to exist in heaven, and is not part of the earthly system with which we are concerned.

^{page 128 note 2} If one interprets 204 cents as a 4 śruti interval and 90 cents as a 2 śruti interval, there would be 24 śrutis in the octave.

^{page 128 note 3} This was in fact stated by Imrat Khan, a surbahār player, in a lecture in London in 1961.

^{page 128 note 4} , Bṛhaddeśī, Trivandrum edition, p. 7, prose following si. 28: ‘As the three registers middle, high, low, i.e. chest, throat, head, are each to be divided into 22, a total of 66 (śrutis) will be reached, so say some’.

^{page 128 note 5} An example of this can be found in the January 1960 issue of the Indian musical journal, . G. N. Goswami, on p. 11, gives the actual values of the rāga Toḍī, in which he gives the sa-ga (minor 3rd) interval as 315 cents (slightly larger than the tempered minor 3rd). In the same issue, Urmila Suri, on p. 19, says that the ga in Toḍī is atikomal (very flat), which should mean something flatter than the tempered minor 3rd. Daniélou, op. cit., II, 49, gives 275 cents (converted into cents by the present writer [N. A. J.]) for this same interval.

^{page 129 note 1} Strangways, A. H. Fox, Music of Hindostan, Oxford, Clarendon Press, 1914, 16Google Scholar. It could be argued that Fox Strangways wrote more than 50 years ago, and that the statement quoted above may have been more meaningful at the time of writing than it is now. It occurs to this writer (N. A. J.) that if the Indian musician's chief asset was his musical ear, he should easily have been able to discriminate against the poor intonation of the harmonium, and that it would not have been accepted by North Indian classical musicians.

^{page 129 note 2} The following is a very incomplete list of the singers and their records on which the harmonium can be heard. Obviously this list would be much larger if it were to include tape recordings and reports of live performances. (1) Abdul Karim Khan (78 r.p.m. Columbia VE 5004), (2) Ustād Faiyaz Khan (78 r.p.m. Hindusthan H 249 L), (3) Roshanara Begum (78 r.p.m. Columbia VE 5037), (4) Narayanrao Vyas (78 r.p.m. HMV N 5725), (5) Bhimsen t Joshi (331/3 r.p.m. HMV ECLP 2253), (6) Hirabai Barodekar (45 r.p.m. HMV 7EPE 1205), I- (7) Kesarbai Kherkar (45 r.p.m. HMV 7ERE 1), (8) Vilayat Hussain Khan (45 r.p.m. HMV I 7EPE 1215), (9) Meera Banerjee (45 r.p.m. HMV 7EPE 1001), (10) Basavraj Rajguru (45 r.p.m. HMV 7EPE 1213), (11) A. T. Kanan (45 r.p.m. HMV 7EPE 1002), (12) Gangoobai Hangal (45 r.p.m. HMV 7EPE 1232), (13) Rasoolanbai (45 r.p.m. Columbia SEDE 3301), (14) Master Krishna (78 r.p.m. HMV P 9692), (15) Shamshadbai (78 r.p.m. Columbia GE 1502), (16) Majoodin JKhan (78 r.p.m. Gramophone Concert Record GC-8–12124), (17) Sangitacarya Ramchandra I hot. Vaze (78 r.p.m. Columbia GE 1532).

From this list there should be no doubt that many of the leading North Indian singers over the past 30 or more years have used the harmonium as an accompanying instrument.

^{page 129 note 3} See the quotations from Daniélou, op. cit., on p. 119 of this article.

^{page 129 note 4} Unfortunately, the present writer (N. A. J.) has not had access to the scales on the harmog niums used by the leading singers. It would no doubt be of great importance to know the exact fe" tunings of these harmoniums.

^{page 130 note 1} There would appear to be a comparable situation among violinists in the West. When accompanied by an instrument with fixed intonation such as the piano, they may be inclined to use the tempered intervals, but when unaccompanied, they may come closer to the ‘natural’ intervals.

^{page 130 note 2} The present writer (N. A. J.) has asked several North Indian musicians to demonstrate śrutis and has usually been given an example in the rāga Darbārī, in which the notes are said to oscillate. In a tape recording of this rāga (in the possession of this writer) played by Ustād Bundu Khan on the , the Ustād interrupts his playing to inform his audience, yeh śruti ki nikhād hai ‘this is the śruti ni’ and proceeds to play it with a wide, slow shake.

^{page 130 note 3} It was Pannalal Ghosh's intonation in this particular recording, with the ga (3rd) sounding too sharp, which was largely responsible for instigating this project. Although, initially the ga in this recording was disturbing, within a minute or two the ear had adjusted itself to this intonation and the melodic line could be appreciated without distraction.

^{page 131 note 1} It may be argued that this conclusion is premature because it is based on the analysis of only two intervals in only one rāga. The rāga Yaman, however, is one of the better known rāgas, and is very often used in teaching the fundamentals of North Indian music, and if there is such a divergence in the intonation of this rāga (each version equally acceptable), there is no reason to expect any greater uniformity in the intonation of other rāgas. It may be mentioned incidentally, that although limited, this investigation has involved several hundred feet of film, and the counting of more than 50,000 waveform peaks.

^{page 131 note 2} Daniélou, Alain, Tableau comparatif des intervalles musicaux, Pondichéry, Institut Français d'Indologie, 1958Google Scholar, introduction, p. iii.

‘Des mesures faites durant de nombreuses années avec des musiciens des diverses régions de l'lnde m'ont permis d'arriver à la conclusion que les intervalles auxquels ceux-ci attribuent des valeure expressives données sont des intervalles précis et déterminés, correspondent toujours aux mêmes valeurs sur les instruments de mesure.’

^{page 132 note 1} Danielou, Alain, Northern Indian music, I, 60Google Scholar.